<counter-nitpick>
The definition of supernatural is not "unexplained". If tomorrow scientists discover there is (say) a fundamental "psychic field" which can only interact directly with human brains and allows direct transmission of thoughts with proper training, that would (if you ask me) fully deserve to be called supernatural, no matter how well explained it is.
To the extent that naturalism is a coherent world-view, the distinction between natural and supernatural has to be different from explained vs unexplained or following natural laws vs not following them. Otherwise, naturalism becomes either trivially false or trivially true.</counter-nitpick>
<Counter-counter-nitpick>
I prefer the term "paranormal" to that. Paranormal is perfectly accurate, if not very precise. "Supernatural" implies that it exists beyond nature, which if it can be detected, that's sort of nonsensical. The difference between naturalism & spiritualism isn't "explained vs. unexplained," so much as "do you have an opinion on things outside of nature." If something IS truly supernatural, there shouldn't be anything to explain. If you have a near death experience, precognition, the ability to see ghosts, etc. you have something natural, there just may or may not be more to it than "bullshit or hallucinations." It doesn't really make sense to call your hypothetical psychic field an example of the supernatural just because that's what it would be considered
today. If it can be empirically verified, why draw that arbitrary distinction? What makes it different from calling a modern electrical system magic, no matter how well it can be explained? That's what it would have been considered at one time.
In other words, naturalism should always do what it was made to do, which is to study nature. If something can be studied in nature, it is natural. This makes the supernatural inherently unfalsifiable, but hey, that's because most who claim it
want it to be that way.
Another way to counter naturalism is the illusion of reality, a la Descartes, so it's not as though there's no theoretical counterpart to it.</counter-counter-nitpick>